CARS Microscopy as an In Situ Characterization Technique for Living Composites

Tissue mimicking materials provide a physiologically relevant 3D environment for cell growth and tissue morphogenesis, shaped in a dynamic reciprocal interaction between the cells and host material. Hence, the matrix must be characterized together with the living cellular component as an integrated system. Traditional cell characterization techniques such as immunofluorescence microscopy are insufficient here, as they probe cell responses without correlating them to the characteristics of the matrix environment. Furthermore, the limited penetration depth of the UV/visible light probe beams and artifacts introduced by fluorophores make it to difficult to characterize 3D cell-matrix organizations and interaction mechanisms. Instead, multi-parametric techniques are required, which can characterize the host material simultaneously with the cells in 3D without physical sectioning. For this purpose, we have developed a robust imaging platform that incorporates spectral coherent anti-Stokes Raman scattering (CARS) microscopy on a standard confocal microscope. CARS microscopy is a label-free method that maps lipids and proteins in cells and the surrounding matrix by probing their inherent molecular vibrations. Through analysis of a broad range of vibrational signatures, resulting in spatially resolved spectral images, we simultaneously characterize the matrix meso-structure and the phenotypes of the resident cells. We have used this technique to study a unique class of recombinant biopolymer hydrogels, termed HELP gels, composed of a recombinant polysaccharide (<b><u>h</u></b>yaluronic acid) and a recombinant <b><u>e</u></b>lastin-<b><u>l</u></b>ike <b><u>p</u></b>rotein (ELP) crosslinked via dynamic covalent chemistry. HELP gels can encapsulate a variety of embedded cells, which alter their morphology and phenotype in response to matrix stiffness. Through CARS analysis of the polymer phase-separation within these hydrogels, we gained a mechanistic understanding of how to alter polymer hydrophilicity to fine-tune hydrogel structure and produce gels with specific mechanical stiffnesses. The mechanical tunability of these hydrogels makes them an attractive platform for developing <i>in vitro</i> models of human tissues and their diseased states.<br/> Ongoing work is exploring these tunable HELP gels as brain mimetic matrices, leveraging CARS to simultaneously probe the dynamic meso-structure of the host gel and the phenotypes of the encapsulated cells based on their lipid profiles. HELP gels were formulated to match the stiffness of brain tissue and used to encapsulate human induced pluripotent stem cell-derived microglia-like cells (iMGs). Microglia are the resident immune cells of the brain, the activation of which is associated with a shift in metabolism and the accumulation of lipid stores. Cell genotypes characterized by metabolic dysfunction have been correlated with the progression of Alzheimer’s disease (AD). By studying the lipid metabolism in HELP-encapsulated iMGs derived from AD-risk patients, detailed insights into the underlying mechanisms will be gained. We show that iMGs are viable in HELP gels and—in the manner of native brain microglia—accumulate lipid droplets, which are detected <i>in situ</i> using CARS microscopy given the unique vibrational signature of lipids. Furthermore, by encapsulating neurons and astrocytes in addition to microglia, we may evaluate lipid transport and metabolism between these cell types. To validate this <i>in vitro</i> model, samples of human brain tissue from healthy donors and AD patients were characterized by CARS microscopy and traditional immunofluorescence microscopy. Consistent with reports from others, we quantified a significant increase in lipid accumulation within the microglia of AD patients (p = 0.05). Taken together, these data demonstrate the usefulness of CARS microscopy to assist in the design, validation, and experimentation of tissue mimics as models of human disease progression.